空中机器人多模态控制研究

On Multi-model Control of Aerial Robotics

研究空中机器人(小型无人直升机)的约束优化控制问题。对小型无人直升机的非线性系统模型进行系统约简,建立混合系统的分段仿射系统模型,针对该问题求解中遇到的在线计算量大,不利于实时控制等问题,提出利用多参数二次规划离线计算出混合系统最优控制律的方法。基于多参数二次规划的方法,在最大可控不变集的可行域内进行显式优化控制器设计,通过反向动态规划,求出对应每一步的优化解,从而求得不变集作为可行域的优化解。通过实际参数的系统仿真,证明了方法的有效性。

The constrained optimal control problem of aerial robotics is studied.The nonlinear system model of small unmanned helicopter is abstracted,and the piece-wise affine hybrid system model is presented.To the complexity of on-line comupation,an explicit optimal controller for hybrid systems based on multi-parametric quadratic programming（mp-QP） is proposed.The feasible domain which is the maximal controlled invariant sets is partitioned in backward dynamic programming by mp-QP method.At each step,the one-step reachable sets are computed,the optimal control laws are constructed to the corresponding regions,and the explicit optimal controller is abtained.Simulation results show the effectiveness of the proposed method.